Anthropic released an update to its Claude platform on October 9, 2026, announcing the ability to orchestrate up to 1,000 AI agents in parallel through dynamic workflows. It is the company's first major update to its managed agents since the initial launch of the feature.
The update enables a lead agent to create a plan, distribute tasks to sub-agents, and merge results upon completion. Up to 1,000 agents can run in parallel per execution, according to the company.
Claude's new dynamic workflows are built on its managed agent infrastructure and are targeted at complex problem-solving tasks requiring coordination among multiple AI agents. Availability begins with the release of the "multiagent_20261001" agent type, initially for developers and enterprise users.
"Whether this is cost-effective seems debatable," said a senior OpenAI engineer. The engineer added that agent swarms can be a massive waste of tokens, but Anthropic points to its own testing.
The announcement follows Anthropic's ongoing efforts to enhance its AI capabilities. The company emphasized that the new feature is part of its broader strategy to improve AI collaboration and efficiency.
Anthropic did not say how the gains from dynamic workflows will hold across different task types, and whether those gains hold across different task types remains to be seen. The company recommended starting small and testing with one's own workloads.
Source: thedecoder
